Yesterday both manager of the year awards were handed out. Coincidently I was reading BP's Between the Numbers book just before the news broke, specifically the chapter written by James Click about evaluating Joe Torre's career. BP attempted to judge managers using the things they had control over; bullpen usage, playing time, and in-game strategy such as intentional bases on balls, sacrifice hits, and stolen base attempts. This sparked my idea on how to judge managers.
Before I begin explaining the method and give the results I want to strongly disclaim these statistics are hardly perfect (in fact the entire thing is quite flawed) however I'm presenting this is a "general idea" rather than a "definite" or "near definite". There are other aspects of managing that I'm clearly ignoring, and you will have to fill those blanks in, unless I go through each team and study the situations and playing time intently. The aspect I will focus on is the last part, IBBs, SHs, ect.
After collecting the mass numbers I ran linear weights on them, with one exception; turning the IBB value into a negative, because remember I'm not weighing team B walking team A's hitters, but rather team A's manager walking team B's hitters. That leaves us with the following formula:
(SBs*0.193)+(CS*-0.282)+(SH*-0.09)+(IBB*-0.176) = Managerial runs created.
TEAM | SB | CS | SH | IBB | lwSBA | lwSH | lwIBB | LWTS | Wins |
Boston | 120 | 35 | 28 | 17 | 13.29 | -2.52 | -2.992 | 7.778 | 0.7778 |
Tampa Bay | 142 | 50 | 23 | 29 | 13.306 | -2.07 | -5.104 | 6.132 | 0.6132 |
LA Angels | 129 | 48 | 32 | 32 | 11.361 | -2.88 | -5.632 | 2.849 | 0.2849 |
NYA | 118 | 39 | 31 | 37 | 11.776 | -2.79 | -6.512 | 2.474 | 0.2474 |
Philadelphia | 136 | 25 | 71 | 64 | 19.198 | -6.39 | -11.264 | 1.544 | 0.1544 |
NY Mets | 138 | 36 | 73 | 53 | 16.482 | -6.57 | -9.328 | 0.584 | 0.0584 |
Oakland | 88 | 21 | 30 | 45 | 11.062 | -2.7 | -7.92 | 0.442 | 0.0442 |
Colorado | 141 | 37 | 90 | 49 | 16.779 | -8.1 | -8.624 | 0.055 | 0.0055 |
Milwaukee | 108 | 38 | 54 | 32 | 10.128 | -4.86 | -5.632 | -0.364 | -0.0364 |
Kansas City | 79 | 38 | 32 | 15 | 4.531 | -2.88 | -2.64 | -0.989 | -0.0989 |
Seattle | 90 | 32 | 36 | 37 | 8.346 | -3.24 | -6.512 | -1.406 | -0.1406 |
Cleveland | 77 | 29 | 43 | 28 | 6.683 | -3.87 | -4.928 | -2.115 | -0.2115 |
Texas | 81 | 25 | 37 | 44 | 8.583 | -3.33 | -7.744 | -2.491 | -0.2491 |
Minnesota | 102 | 42 | 52 | 38 | 7.842 | -4.68 | -6.688 | -3.526 | -0.3526 |
LA Dodgers | 126 | 43 | 64 | 58 | 12.192 | -5.76 | -10.208 | -3.776 | -0.3776 |
Toronto | 80 | 27 | 48 | 42 | 7.826 | -4.32 | -7.392 | -3.886 | -0.3886 |
Baltimore | 81 | 37 | 27 | 44 | 5.199 | -2.43 | -7.744 | -4.975 | -0.4975 |
St. Louis | 73 | 32 | 71 | 21 | 5.065 | -6.39 | -3.696 | -5.021 | -0.5021 |
Pittsburgh | 57 | 19 | 66 | 31 | 5.643 | -5.94 | -5.456 | -5.753 | -0.5753 |
ChicagoN | 87 | 34 | 65 | 45 | 7.203 | -5.85 | -7.92 | -6.567 | -0.6567 |
ChicagoA | 67 | 34 | 28 | 42 | 3.343 | -2.52 | -7.392 | -6.569 | -0.6569 |
Houston | 114 | 52 | 57 | 53 | 7.338 | -5.13 | -9.328 | -7.12 | -0.712 |
San Fran | 108 | 46 | 57 | 59 | 7.872 | -5.13 | -10.384 | -7.642 | -0.7642 |
Arizona | 58 | 23 | 68 | 41 | 4.708 | -6.12 | -7.216 | -8.628 | -0.8628 |
Florida | 76 | 28 | 49 | 66 | 6.772 | -4.41 | -11.616 | -9.254 | -0.9254 |
Washington | 81 | 43 | 64 | 44 | 3.507 | -5.76 | -7.744 | -9.997 | -0.9997 |
Cincinnati | 85 | 47 | 72 | 40 | 3.151 | -6.48 | -7.04 | -10.369 | -1.0369 |
Detroit | 63 | 31 | 30 | 63 | 3.417 | -2.7 | -11.088 | -10.371 | -1.0371 |
Atlanta | 58 | 27 | 69 | 61 | 3.58 | -6.21 | -10.736 | -13.366 | -1.3366 |
San Diego | 36 | 17 | 59 | 61 | 2.154 | -5.31 | -10.736 | -13.892 | -1.3892 |
After I accumulated the data, I talked with Peter and Sky about actually using it. My biggest concern (as well as theirs) was the lack of context in these numbers. The abundance of negative numbers is to be expected, after all the only way managers can "gain points" in this measurement is by having a successful stealing team, and even that is part luck. After some more discussion, I arrived at this "solution": using Baseball-Reference's "high leverage" splits for pitching and offense to compile the numbers. I did so, and let me just inform everyone I now know every team B-Ref's tag, fun!
Here are the numbers in high leverage situations:
TEAM | SB | CS | SH | IBB | lwSBA | lwSH | lwIBB | LWTS | Wins |
Boston | 32 | 6 | 14 | 9 | 4.484 | -1.26 | -1.584 | 1.64 | 0.164 |
NYA | 38 | 8 | 16 | 19 | 5.078 | -1.44 | -3.344 | 0.294 | 0.0294 |
Seattle | 29 | 7 | 16 | 12 | 3.623 | -1.44 | -2.112 | 0.071 | 0.0071 |
Philadelphia | 42 | 6 | 30 | 22 | 6.414 | -2.7 | -3.872 | -0.158 | -0.0158 |
NY Mets | 42 | 4 | 34 | 25 | 6.978 | -3.06 | -4.4 | -0.482 | -0.0482 |
ChicagoA | 21 | 8 | 13 | 8 | 1.797 | -1.17 | -1.408 | -0.781 | -0.0781 |
Minnesota | 23 | 10 | 8 | 11 | 1.619 | -0.72 | -1.936 | -1.037 | -0.1037 |
Tampa Bay | 44 | 21 | 9 | 16 | 2.57 | -0.81 | -2.816 | -1.056 | -0.1056 |
Houston | 35 | 13 | 26 | 14 | 3.089 | -2.34 | -2.464 | -1.715 | -0.1715 |
Milwaukee | 29 | 9 | 30 | 13 | 3.059 | -2.7 | -2.288 | -1.929 | -0.1929 |
LA Dodgers | 35 | 10 | 21 | 26 | 3.935 | -1.89 | -4.576 | -2.531 | -0.2531 |
St. Louis | 18 | 10 | 16 | 10 | 0.654 | -1.44 | -1.76 | -2.546 | -0.2546 |
Baltimore | 26 | 10 | 12 | 21 | 2.198 | -1.08 | -3.696 | -2.578 | -0.2578 |
LA Angels | 34 | 13 | 28 | 17 | 2.896 | -2.52 | -2.992 | -2.616 | -0.2616 |
Pittsburgh | 25 | 4 | 39 | 16 | 3.697 | -3.51 | -2.816 | -2.629 | -0.2629 |
Arizona | 18 | 5 | 34 | 10 | 2.064 | -3.06 | -1.76 | -2.756 | -0.2756 |
Cincinnati | 31 | 12 | 32 | 15 | 2.599 | -2.88 | -2.64 | -2.921 | -0.2921 |
Kansas City | 17 | 10 | 24 | 7 | 0.461 | -2.16 | -1.232 | -2.931 | -0.2931 |
Cleveland | 18 | 7 | 29 | 11 | 1.5 | -2.61 | -1.936 | -3.046 | -0.3046 |
Oakland | 24 | 8 | 20 | 23 | 2.376 | -1.8 | -4.048 | -3.472 | -0.3472 |
Florida | 21 | 6 | 24 | 22 | 2.361 | -2.16 | -3.872 | -3.671 | -0.3671 |
Colorado | 37 | 16 | 43 | 14 | 2.629 | -3.87 | -2.464 | -3.705 | -0.3705 |
Toronto | 21 | 4 | 30 | 23 | 2.925 | -2.7 | -4.048 | -3.823 | -0.3823 |
San Fran | 45 | 18 | 32 | 26 | 3.609 | -2.88 | -4.576 | -3.847 | -0.3847 |
Texas | 23 | 10 | 21 | 22 | 1.619 | -1.89 | -3.872 | -4.143 | -0.4143 |
ChicagoN | 31 | 11 | 31 | 29 | 2.881 | -2.79 | -5.104 | -5.013 | -0.5013 |
Atlanta | 17 | 7 | 28 | 23 | 1.307 | -2.52 | -4.048 | -5.261 | -0.5261 |
San Diego | 14 | 7 | 28 | 20 | 0.728 | -2.52 | -3.52 | -5.312 | -0.5312 |
Detroit | 11 | 9 | 22 | 23 | -0.415 | -1.98 | -4.048 | -6.443 | -0.6443 |
Washington | 22 | 15 | 32 | 25 | 0.016 | -2.88 | -4.4 | -7.264 | -0.7264 |
Boston again rates at the top. Many broadcasters talk Terry Francona up as a great manager, but for all the wrong reasons. I am sure he is a nice person, and the players like him, yet the guy is a pretty good strategist if you believe these numbers. It doesn't hurt having great personnel and an outstanding front office either. Joe Girardi's high leverage non-personnel tactics are absolutely fine. Amusingly, so were the combination of John McLaren and Jim Riggleman. That's one of those situations where the personnel is the issue. Atlanta, San Diego, Detroit, and Washington again rank low, perhaps Bobby Cox's best is behind him?
Interestingly both of the World Series teams finish within the top 10 on both scales, but Joe Maddon becomes far worse in high leverage situations, unfortunately his worst habits seemed to follow him into the World Series, as MGL covered nicely here. Even still Maddon's regular season methods fell secondary to Francona's in both instances, although his team did win more games, but you know how processes are more important than results in the long-term.
Here's a look at the averages:
Team | NonLvg | Lvg | Avg |
San Diego | 30 | 28 | 29 |
Detroit | 28 | 29 | 28.5 |
Atlanta | 29 | 27 | 28 |
Washington | 26 | 30 | 28 |
San Fran | 23 | 24 | 23.5 |
Florida | 25 | 21 | 23 |
ChicagoN | 20 | 26 | 23 |
Cincinnati | 27 | 17 | 22 |
Arizona | 24 | 16 | 20 |
Toronto | 16 | 23 | 19.5 |
Texas | 13 | 25 | 19 |
Pittsburgh | 19 | 15 | 17 |
Houston | 22 | 9 | 15.5 |
Cleveland | 12 | 19 | 15.5 |
St. Louis | 18 | 12 | 15 |
Baltimore | 17 | 13 | 15 |
Colorado | 8 | 22 | 15 |
Kansas City | 10 | 18 | 14 |
ChicagoA | 21 | 6 | 13.5 |
Oakland | 7 | 20 | 13.5 |
LA Dodgers | 15 | 11 | 13 |
Minnesota | 14 | 7 | 10.5 |
Milwaukee | 9 | 10 | 9.5 |
LA Angels | 3 | 14 | 8.5 |
Seattle | 11 | 3 | 7 |
NY Mets | 6 | 5 | 5.5 |
Tampa Bay | 2 | 8 | 5 |
Philadelphia | 5 | 4 | 4.5 |
NYA | 4 | 2 | 3 |
Boston | 1 | 1 | 1 |
It's tricky using these as is, but only a single seasons worth of data likely isn't worth much either. That's why I'm disclaiming the heck out of this and hoping people take it with a grain of salt, it's only a part of what managers do, and it's a rough attempt at quantifying that part.